Eighteenth Australasian Weeds Conference

Putting fireweed on the front burner: improving management and understanding impact

Brian M. Sindel1, Michael J. Coleman1 and Ian J. Reeve2 1 School of Environmental and Rural Science, University of New England, Armidale NSW 2351 2 Institute for Rural Futures, University of New England, Armidale NSW 2351 ([email protected])

Summary Fireweed (Senecio madagascariensis) Significance (AWC 2012). A joint University of New continues to spread in coastal pastures in south eastern England and CSIRO research project sought to address , as well as onto the Northern and Southern gaps in knowledge relating to fireweed ecology, impact Tablelands of NSW and in South-East Queensland. and management, through field trials and survey of This paper details the key findings of a national sur- landholders, and commenced research into biologi- vey of landholders conducted in late 2011 to evaluate cal control options in the native range for fireweed in fireweed impact and management options. South . In this paper we focus on the landholder Fireweed was less likely to be considered a prob- survey findings, and the implications for improving lem in areas where it had spread recently, or where it fireweed management. had a long history of infestation. It was considered the ‘worst weed on property’ by about one third of MATERIALS AND METHODS respondents. Over 80% of respondents attempted to Survey design The questionnaire form was based on control fireweed. Many use a mixture of methods, an earlier fireweed survey conducted in 1985 (Sindel amongst which the more successful appeared to be 1989). Nearly all questions from the original survey hand weeding, grazing with sheep or goats, herbicide were included unchanged in the new form to be able and promoting competitive pasture. The most sig- to observe changes that may have occurred in the nificant economic impacts of fireweed included lack fireweed situation over the last 25 years or so. New of time to devote to other farm tasks, and impact on questions were also added. The draft questionnaire farm profitability. Even landholders who considered was evaluated by project steering committee members, fireweed under control on their farm spent consider- landholders, and NSW and Queensland (Qld) govern- able time and money managing the weed. Fireweed ment staff, to ensure the questionnaire addressed the remains a considerable concern for landholders along most topical issues. the south-eastern Australian coast and hinterland. Keywords Fireweed, spread, occurrence, con- Distribution Current distribution data were used trol methods, impact. to identify post codes representing the known current extent of fireweed spread. As in the 1985 survey, post INTRODUCTION codes were grouped into a number of to allow Fireweed is one of the worst weeds of coastal pastures comparisons between regions, and regional compari- in south-eastern Australia. It contains toxic pyrrolizi- sons between the 1985 and 2011 surveys (Figure 1). dine alkaloids that, if consumed, cause liver damage in Postal distribution focused on dairy and beef graz- livestock, leading in some cases, to death (Walker and ing properties, though other producers were welcome to Kirkland 1981). It is thought to have been introduced contribute. A number of producer groups distributed the to Australia from Southern Africa in shipping ballast, survey to their members. Additional graziers and dairy with the earliest recorded specimen having been col- farmers were identified in the Yellow Pages. A final lected near , NSW, in 1918 (Sindel mailout of 1764 surveys occurred in November 2011. 1996). Fireweed is now widespread in coastal pastures A concurrent internet survey was developed to from South-East Queensland to Bega in Southern supplement the mail survey in the event of a lower NSW, and has established in higher altitude locations than expected response. This survey was promoted such as the Dorrigo, Tenterfield and Nowendoc dis- through a number of producer group and natural tricts in the NSW , and the eastern resource management email lists and e-newsletters. of the . Fireweed is therefore of considerable concern to Analysis Survey respondents were grouped into farmers and rural communities in impacted areas. Its regions based on post code, where a post code was significance was confirmed in April, 2012, when it provided (Figure 1). A number of new regions were was named as one of twelve new Weeds of National included in the 2011 survey, as fireweed had expanded

166 Eighteenth Australasian Weeds Conference its range since 1985. Analysis was con- ducted using SPSS 20 (IBM 2011). Comparative regional analysis be- tween the 1985 and 2011 results was conducted where questions were identical. Statistically significant changes between the two data sets were identified using the binomial test for nominal variables, and the Mann-Whitney U test for ordinal variables.

RESULTS A 30% response rate was obtained from the mail survey (528 responses), with a further 403 responses made to the online survey, comprising a total of 931 respons- es. Not all 931 respondents, however, had fireweed on their property, and so were not required to complete all questions.

Occurrence and spread Fireweed has Figure 1. Regions sampled in the survey 2011. continued to spread into new regions. It appears to be a relatively recent arrival in the NSW Northern and Southern Table- lands, parts of the NSW Mid-North Coast, and South-East Qld. Severe fireweed infestations have become more common in Bega, Shoalhaven, Lismore, NSW Mid-North Coast and South-East Qld. It has been present for longer in parts of the NSW far north coast, around and Wauchope, the Gloucester/ region, the lower Hunter Valley, and the south-west fringe of (Figure 2). Fireweed was less likely to be con- sidered a major problem in some regions where it has only spread most recently (such as the NSW Northern and Southern Tablelands, and South-East Qld), or where it had occurred for the longest (such as Muswellbrook and Gloucester) (Table 1). It may be that respondents from the former were not yet fully impacted by Figure 2. Duration of fireweed presence, mean by post code. fireweed, and those from the latter had learnt to manage the weed more effectively. for reasons including: that it was a losing battle, that Fireweed was the worst weed for about a third control had proven ineffective, that it took too much of respondents, and was more likely to be considered out of farm finances, and that it was not worth doing the worst weed in Bega, Shoalhaven, Hexham and the when neighbours were not acting to manage the weed. Mid-North Coast, and least likely in Muswellbrook Since 1985, slashing and cultivation were signifi- and Gloucester. cantly less likely to be used, while the use of herbicides and grazing with sheep or goats appeared to have Control methods Over 80% of respondents attempt- increased, though the latter result may in part be due ed to control fireweed using one or more methods. to the inclusion of non-beef and dairy landholders in Approximately 12% no longer controlled the weed, the more recent survey (Table 2).

167 Eighteenth Australasian Weeds Conference

Hand weeding remained the most common and The proportion of farmers controlling fireweed one of the most successful control methods, though with herbicide increased significantly, particularly in it is time-consuming and physically demanding. It is the Cumberland, Mid-North Coast, South-East Qld, particularly favoured for targeted control on specific and Shoalhaven regions. Those who had a moderate areas of the property, or smaller outbreaks. Those to large fireweed infestation on their property were respondents from regions which had experienced more likely to favour herbicide use than those who less fireweed pressure were more likely to remove only had a small infestation. it by hand. While promoting competitive pasture is an im- Despite the ongoing relative popularity of slash- portant longer term control method for fireweed, it ing, many respondents suggested it makes the fireweed appeared that the use of this method had not increased problem worse by spreading seed. (Table 2). Where pasture improvement was used suc- cessfully to reduce the intensity of fireweed, healthy pastures were encouraged through application of Table 1. Size of the fireweed problem as perceived fertiliser and lime, promotion of native and introduced by farmers, as a percentage of respondents with pasture species, and reduced grazing pressure or graz- fireweed 2011. ing rotation strategies. No/ Moderate/ Grazing with sheep or goats was considered to be minor major the most successful fireweed control method overall, Region problem problem n though was still only used by a relatively small pro- Overall survey* 52 48 788 portion of respondents. Sheep or goat producers spent Cumberland 65 35 29 less time and money on fireweed control, and were Hexham 49 51 45 less likely to consider fireweed to have an impact Muswellbrook 63 37 35 on farm profitability. However, establishing sheep Northern Tablelands/ 68 32 25 NW Slopes or goats on a cattle property can involve significant Gloucester 65 35 20 infrastructure outlays. Taree/ Wauchope 49 51 70 Those who had successfully reduced the fireweed Mid-North Coast 49 51 80 on their property were most likely to credit their suc- Lismore 43 57 83 cess to one, or a mixture, of hand weeding, pasture Shoalhaven 46 54 106 competition, and herbicide application. Often these Bega 45 55 128 and other approaches were successfully used together Southern inland NSW 73 27 11 as part of an integrated strategy. South-East Qld 62 38 125 Over 50% of respondents considered that a bio- * includes those who did not provide a post code. logical control option for fireweed would be a ‘very chisq = 23.1, df = 12, p = 0.027. important’ development. Those most likely to support biological control appeared to have had fireweed on their property for longer, and larger infestations. Table 2. Use and success of fireweed control meth- Over 60% of respondents believed fireweed ods, as a percentage of respondents with fireweed should be declared noxious or prohibited under leg- 2011 (1985). islation. Those who had fireweed on their farm for Level of longer, or who had larger infestations, often found it Level of success n an intractable farm management issue, and were less Method use* “high”** (2011 only) likely to support declaration or prohibition. Those Hand weeding† 81 (74) 53 (34) 570 facing fireweed incursion for a shorter time, or with Slashing† 36 (68) 13 (13) 274 a smaller infestation, were more likely to support Cultivation† 4 (19) 18 (13) 62 declaration or prohibition. Herbicides† 38 (12) 41 (41) 259 Grazing with sheep 11 (5) 54 (67) 78 or goats† Economic impact The most significant economic Reduced stocking 9 (n/a) 22 (n/a) 82 impacts of fireweed included lack of time to devote rates to other activities, and overall farm profitability. Promoting 35 (35) 32 (42) 234 Nearly 50% of respondents estimated spending over competitive pasture 50 hours each year controlling fireweed, while nearly * % respondents who attempt fireweed control. 40% estimated spending $1000 or more on control ** % respondents using that method. activity. Of those who considered fireweed to be un- † Binomial test p = 0. der control, approximately 50% spend more than 50

168 Eighteenth Australasian Weeds Conference hours and more than $1000 per year. This suggests required to quantify the economic impact of fireweed that keeping fireweed at a manageable level on farm on agricultural production, including identifying the can be a significant ongoing cost to farmers in both practical level of tolerance farmers can assume for time and money. fireweed, benefit/cost analyses of its impact on pasture Fireweed control appeared to be particularly time production, economic impact on a national scale, and consuming in the Shoalhaven, Bega and Mid-North impact on individual farmer livelihoods. Coast regions, and it was notable that these regions As this research has shown, fireweed may be were amongst those where fireweed was more likely managed with some success using a range of control to be considered a moderate or major problem. Nearly methods. However, many landholders expressed dis- 20% of respondents considered that fireweed impacted couragement about their ongoing or future capacity to negatively on livestock health. manage it, considering a successful biological control agent to be of the utmost importance. Biological con- DISCUSSION trol research is therefore a high ongoing priority for Fireweed remains a considerable concern for landhold- fireweed management. ers. It is highly invasive, and has had a significant economic impact in affected regions. ACKNOWLEDGMENTS It is particularly notable that fireweed was often We gratefully acknowledge the funding provided for less likely to be considered a moderate or major this research by the Australian Government Depart- problem in those regions where the weed appeared to ment of Agriculture, Fisheries and Forestry (DAFF). have spread more recently, as well as regions where We appreciate the assistance and support of the firew- the weed had been present for a longer time. In areas eed control research steering committee, and gratefully with a long history of infestation, fireweed spread acknowledge the time taken by all landholders who from one farm to the next has largely ceased (Sindel completed and returned our survey. 2009). In parts of the Bega Valley, Shoalhaven, and Mid-North Coast of NSW, however, the weed appeared REFERENCES to have been present on many properties for between Australian Weeds Committee (AWC). (2012). Weeds 20 and 30 years. It was generally in these areas where of National Significance. http://www.weeds.org. fireweed has caused the greatest concern and continues au/WoNS/, accessed 15/05/12. to spread rapidly. IBM. (2011). SPSS Statistics Version 20. IBM, New Although fireweed can be particularly difficult to York. control for many landholders, this research suggests Launders, T.E. (1986). Competitive pastures and that effective management is possible using an inte- control of fireweed. Australian Weeds Research grated approach that may incorporate hand weeding, Newsletter 35, 42-3. herbicide application, and in some circumstances Madden, J. (2008). Socio economics of fireweed. grazing with sheep or goats. Underpinning a successful Presentation to the National Fireweed Conference, longer term fireweed management strategy on farm Bega RSL (Bega Valley Fireweed Association, will be maintaining a vigorous, competitive pasture Bega) http://fireweed.org.au/16618.html, accessed (Launders 1986). 15/05/12. Landholders need to be made aware through Sindel, B.M. (1989). The ecology and control of extension of how to improve their capacity to success- fireweed (Senecio madagascariensis Poir.). PhD fully manage fireweed. Of particular importance will Thesis, University of Sydney. be extending best practice control to those landholders Sindel, B.M. (1996). Impact, ecology and control of who currently show little interest in controlling firew- the weed Senecio madagascariensis in Australia. eed. Extension must highlight the need for vigilance Proceedings of the International Compositae Con- and long-term dedication in monitoring and manage- ference, Vol. 2 Biology and Utilization, (eds) P. ment of fireweed, as many respondents suggested Caligari and D.J.N. Hind, pp. 339-49. (Royal these were important elements behind their success. Botanic Gardens, Kew). Fireweed appears to have a significant economic Sindel, B.M. (2009). Fireweed in Australia: directions impact on farmers, including those who are managing for future research. Report to the Bega Valley the weed with some success. Recent research con- Fireweed Association, Bega. http://fireweed.org. servatively estimated that fireweed had the capacity au/15829.html, accessed 15/05/12. to reduce the output of broadacre, low input grazing Walker, K.H. and Kirkland, P.D. (1981). Senecio systems in the South Coast region of lautus toxicity in cattle. Australian Veterinary by 20% (Madden 2008). More detailed information is Journal 57, 1-7.

169